OT/FOAK: Multiple PSUs in a PC

Got a problem with the PC, and I suspect it may be due to the power supply
not being meaty enough.

Problem: The PC just randomly reboots, no crash, no BSOD, just suddenly off,
and then back to the BIOS screen. When I get back into Windows, there's no
trace of a problem in the event log, no "Your PC has recovered from a
serious error" etc. It's as if it didn't happen, apart from the losing the
work I was doing :-( Perhaps coincidentally, it seems to reboot more often
when running a BitTorrent client (either G3Torrent or Azureus).

I'm running 8 hard discs and 2 DVD drives, as well as a dualhead graphics
card. It could be that this is stressing the PSU somewhat, but I already
have a 550W PSU in. I have room in the tower case for a second PSU, but I
don't know how to get it to work, as they are controlled via the motherboard
these days aren't they?

Alternatively, could having too many open connections with bittorrent be
enough to cause the PC to bomb out?

Suggestions/comments/abuse etc welcomed.

 

I just binned a machine that was doing that. I did however get "recovered
from a serious error" messages although that was the extent of their
feedback. Mine would just reboot whenever I asked it to do some slightly
serious work, like anything that wasn't just sitting there. It became more
and more frequent. I ran memtest on it which threw up a load of errors but I
had no similar sized ram to exchange and swapping the sticks around didn't
produce any positive result. Being over 3 years old, I whipped all the bits
out and chucked it. The bits will remain in a box under the desk until I
eventuially throw them out as well.

 

Are you using a Netgear FA311 network card in all of this? I ask as a
friend of mine had a similar problem, especially when he was using
BitTorrent. After reading up on a few potential problems with this
particular card, I replaced it with a manky old 3Com one I had lurking
around and all is now well.

As for linking up 2 PSUs, it can be done with a bit of trickery.

http://www.burningissues.net/how_to/power/psu.htm


Failing that, get an old mains-switched AT supply and use that.

 

Certainly sounds like it might be power related.

Yes, if you've got a crap card / driver.


--
Catman MIB#14 SKoGA#6 TEAR#4 BOTAFOF#38 Apostle#21 COSOC#3
Tyger, Tyger Burning Bright (Remove rust to reply)
Alfa 116 Giulietta 3.0l (Really) Sprint 1.7 75 2.0 TS
Triumph Speed Triple: Black with extra black bits
www.cuore-sportivo.co.uk

 

Actually, it's one of the two network ports built into the motherboard (Asus
AN78E-Deluxe V2)

<checks>

Ok, the one that's connected currently is a "3Com 32920B-EMB Integrated fast
ethernet controller"

The other one, nothing plugged in is an "NVidia nForce Networking
Controller"

Ironically, I have a Netgear FA-311 card sat in my box of spares!

Ta, will look into that.

Good point. I only want it to drive half the hard disks after all.

 

PC actually crashed whilst I was writing this reply first time!

Ok, there is no s/w firewall on board, I leave that to the router. Swept
every day for spyware and viruses so, t'aint that.

CPU temp - 42C
Mobo temp - 26C
+12V = 12.608
+5V = 4.999
+3.3V = 3.296
VCore = 1.712

Don't have a way of getting GPU temp, but I doubt it's that. It's happens
whatever I'm doing, whether it involves intensive CPU/GPU work or not.

Hard disk temps... Hmmm, some of them *are* in acoustic enclosures, but they
do have pretty hefty heatsinks on them.

I'm actually leaning towards the idea of it being excessive network
connections due to bitTorrent client running, but I haven't done a memory
test yet.

 

There's your problem.

 

Mr. Fantastic wrote

Yeah, kin amazing innit eh. I have got this old piece of shite sitting
on the table here and the case is glowing red where the fan can't shit
the air through the dust and filth inside and it works a treat. The
minute I stuff a thermometer through the little hole on the font it
stops working.

 

Really? The warning alarm in the BIOS is preset at 70C for the processor. I
was under the impression that anything below about 60 degrees wouldn't
bother it in the slightest.

 

It was somewhere around Barstow, on the edge of the desert, when the
drugs began to take hold. I remember "Mr. Fantastic"

Both of those are well within limits.

--

Dave

GS 850 x2 / SE 6a
SbS#6 DIAABTCOD#16 APOSTLE#6 FUB#3
FUB KotL OSOS#12? UKRMMA#19 COSOC#10

 

I think he was looking at the '-' sign...
--
Catman MIB#14 SKoGA#6 TEAR#4 BOTAFOF#38 Apostle#21 COSOC#3
Tyger, Tyger Burning Bright (Remove rust to reply)
Alfa 116 Giulietta 3.0l (Really) Sprint 1.7 75 2.0 TS
Triumph Speed Triple: Black with extra black bits
www.cuore-sportivo.co.uk

 

Could be ropey APIC handling between BIOS and O/S although not a problem
I've experienced under XP, only certain flavours of Linux.
--
| Wik -UKRMHRC#10- 2003 R1150GSA -DC#1 -'FOT#0 'FOF #39 - BOD#12 BOB#12
|# You don't believe me | "Experience is the worst teacher.
|That the scenery | It always gives the test first
|Could be a cold-blooded killer. | and the instruction afterward."
***** human response from wik at blueyonder dot co dot uk *****

 

Intel or AMD? If it's Intel it's "warm" and unless it's o/c'd, I'd be
looking to see if the heat-sink/fan is doing 'owt.

If it's an Athlon then it's just warmimg up,

I run Intel exclusively (don't ask me why) and I'm running a bone stock
P4HT 3.0GHz with the standard Intel 'sink'n'fan combo and it's running
at <checks>:
Field Value
Sensor Properties

Temperatures
Motherboard 27 °C (81 °F)
CPU 24 °C (75 °F)
Maxtor 6Y160M0 48 °C (118 °F)

Cooling Fans
CPU 1896 RPM
Power Supply 1406 RPM

Check your H/D temps, too. Everest Home edition is your friend, here or
the "sensors" kernel module for Linux (assuming they're S.M.A.R.T. HDDs
made within the last 5 years or so).

60°C is about cooking temp for the average 3.5" IDE drive and mine, as
you can see, is at nearly 50 without any acoustic enclosures...

As an aside, as someone mentioned GPU temps, my GeForce 6800 has a
slowdown threshold of a staggering 135°C!
=8-0

--
| Wik -UKRMHRC#10- 2003 R1150GSA -DC#1 -'FOT#0 'FOF #39 - BOD#12 BOB#12
|# You don't believe me | "Experience is the worst teacher.
|That the scenery | It always gives the test first
|Could be a cold-blooded killer. | and the instruction afterward."
***** human response from wik at blueyonder dot co dot uk *****

 

It's an Athlon all right!

They are, and I will. Cheers for that.

Yeah, I managed to kill a 180Gb hard drive that way[1]. That's why only
*some* are in enclosures now


[1] Bizarrely it worked after the "stick it in a freezer for a few hours"
trick. Long enough for me to copy the data off, wipe it then get a new one
under warranty ;-)

 

Wireless network by any chance? My laptop is guaranteed to BSOD if I
run p2p heavily over wireless but not when physically connected.
Intel Pro Wireless 2200b/g.

 

*whooosh*

 

It was somewhere around Barstow, on the edge of the desert, when the
drugs began to take hold. I remember "Mr. Fantastic"

Whooosh yersel'.

--

Dave

GS 850 x2 / SE 6a
SbS#6 DIAABTCOD#16 APOSTLE#6 FUB#3
FUB KotL OSOS#12? UKRMMA#19 COSOC#10

 

Aaah, the classic I've fucked so'll claim a "whoosh". Good call.

 

Much loved of Ace and Cab....

 

On Fri, 28 Jan 2005 15:25:09 +0000, Dan White wrote:

[snip PC woes]

Bad RAM.
Try running burnintestpro from Passmark software. 30 day trial version.